How do wartime casualties affect support for the incumbent in authoritarian regimes? Drawing from observational data on casualties and voting, as well as a survey experiment, this study provides empirical evidence from Russia after the full-scale invasion of Ukraine.

In an observational study pre-registered before the 2024 presidential elections, we test several literature-driven hypotheses on how casualties affect voting for the incumbent. The study employs local-level data on casualties from open sources collected by independent journalists and volunteers. We employ advanced difference-in-difference analyses to assess the relationships between exposure to casualties and electoral support.

In a pre-registered survey experiment built into the 2024 Russian Election Study of about 1600 adult individuals, we randomize exposure to an indirect benchmark of casualties in large-scale conflicts. This vignette experiment provides all respondents with short pieces of information about the current Russo-Ukrainian war. While all respondents receive information that the number of Russian soldiers in Ukraine is very close to the number of Soviet soldiers in Afghanistan, we are randomly assigning some respondents to receive additional information about casualties from the Soviet-Afghan War. By providing an indirect reference, the experiment is designed to elicit the effect of exposure to casualty information on support for the incumbent. The survey is currently in the field, as of December 2024. Together with the observational data analysis, the survey experiment will provide a more robust and internally valid assessment of casualty sensitivity. This study will offer crucial insights into public accountability and regime support in authoritarian contexts.